spotify_classifier_new method_large_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import scipy as sp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1870455\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>session_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>skip_1</th>\n",
       "      <th>skip_2</th>\n",
       "      <th>skip_3</th>\n",
       "      <th>not_skipped</th>\n",
       "      <th>context_switch</th>\n",
       "      <th>hist_user_behavior_reason_start</th>\n",
       "      <th>hist_user_behavior_reason_end</th>\n",
       "      <th>hist_user_behavior_is_shuffle</th>\n",
       "      <th>skip_score</th>\n",
       "      <th>liking</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0_00001e52-55c0-470c-8d32-85c691c2019a</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.3</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0_00a20441-f168-41c7-9bf6-93ddeb022f00</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0_00e25cfa-fb17-4c17-a126-04dc55c66259</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0_05c168a6-d905-427e-8a6a-4b12311fcb6d</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.3</td>\n",
       "      <td>2.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0_05d6ac38-9e98-44de-b097-048f1c0f95db</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                              session_id  \\\n",
       "0           0  0_00001e52-55c0-470c-8d32-85c691c2019a   \n",
       "1           1  0_00a20441-f168-41c7-9bf6-93ddeb022f00   \n",
       "2           2  0_00e25cfa-fb17-4c17-a126-04dc55c66259   \n",
       "3           3  0_05c168a6-d905-427e-8a6a-4b12311fcb6d   \n",
       "4           4  0_05d6ac38-9e98-44de-b097-048f1c0f95db   \n",
       "\n",
       "                                 track_id  skip_1  skip_2  skip_3  \\\n",
       "0  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0    -0.5     0.2   \n",
       "1  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.2   \n",
       "2  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.2   \n",
       "3  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.0   \n",
       "4  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.0   \n",
       "\n",
       "   not_skipped  context_switch  hist_user_behavior_reason_start  \\\n",
       "0            0               0                              0.2   \n",
       "1            0               0                              0.2   \n",
       "2            0               0                              0.2   \n",
       "3            1               0                              0.3   \n",
       "4            1               0                              1.0   \n",
       "\n",
       "   hist_user_behavior_reason_end  hist_user_behavior_is_shuffle  skip_score  \\\n",
       "0                            0.4                              0        -0.3   \n",
       "1                            0.4                              0         0.2   \n",
       "2                            0.4                              0         0.2   \n",
       "3                            1.0                              0         1.0   \n",
       "4                            1.0                              0         1.0   \n",
       "\n",
       "   liking  rating  \n",
       "0     0.6     0.3  \n",
       "1     0.6     0.8  \n",
       "2     0.6     0.8  \n",
       "3     1.3     2.3  \n",
       "4     2.0     3.0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(r'merged.csv')\n",
    "print(len(df))\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    1.870440e+06\n",
       "mean     1.257080e+00\n",
       "std      1.475866e+00\n",
       "min     -1.300000e+00\n",
       "25%      3.000000e-01\n",
       "50%      1.000000e+00\n",
       "75%      2.600000e+00\n",
       "max      4.000000e+00\n",
       "Name: rating, dtype: float64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['rating'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['rating_round'] = df['rating'].round()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>session_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>skip_1</th>\n",
       "      <th>skip_2</th>\n",
       "      <th>skip_3</th>\n",
       "      <th>not_skipped</th>\n",
       "      <th>context_switch</th>\n",
       "      <th>hist_user_behavior_reason_start</th>\n",
       "      <th>hist_user_behavior_reason_end</th>\n",
       "      <th>hist_user_behavior_is_shuffle</th>\n",
       "      <th>skip_score</th>\n",
       "      <th>liking</th>\n",
       "      <th>rating</th>\n",
       "      <th>rating_round</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0_00001e52-55c0-470c-8d32-85c691c2019a</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.3</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.3</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0_00a20441-f168-41c7-9bf6-93ddeb022f00</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.8</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0_00e25cfa-fb17-4c17-a126-04dc55c66259</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.8</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0_05c168a6-d905-427e-8a6a-4b12311fcb6d</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.3</td>\n",
       "      <td>2.3</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0_05d6ac38-9e98-44de-b097-048f1c0f95db</td>\n",
       "      <td>t_a0b2a533-caad-4a72-af14-e647bd7e9c1f</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                              session_id  \\\n",
       "0           0  0_00001e52-55c0-470c-8d32-85c691c2019a   \n",
       "1           1  0_00a20441-f168-41c7-9bf6-93ddeb022f00   \n",
       "2           2  0_00e25cfa-fb17-4c17-a126-04dc55c66259   \n",
       "3           3  0_05c168a6-d905-427e-8a6a-4b12311fcb6d   \n",
       "4           4  0_05d6ac38-9e98-44de-b097-048f1c0f95db   \n",
       "\n",
       "                                 track_id  skip_1  skip_2  skip_3  \\\n",
       "0  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0    -0.5     0.2   \n",
       "1  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.2   \n",
       "2  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.2   \n",
       "3  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.0   \n",
       "4  t_a0b2a533-caad-4a72-af14-e647bd7e9c1f       0     0.0     0.0   \n",
       "\n",
       "   not_skipped  context_switch  hist_user_behavior_reason_start  \\\n",
       "0            0               0                              0.2   \n",
       "1            0               0                              0.2   \n",
       "2            0               0                              0.2   \n",
       "3            1               0                              0.3   \n",
       "4            1               0                              1.0   \n",
       "\n",
       "   hist_user_behavior_reason_end  hist_user_behavior_is_shuffle  skip_score  \\\n",
       "0                            0.4                              0        -0.3   \n",
       "1                            0.4                              0         0.2   \n",
       "2                            0.4                              0         0.2   \n",
       "3                            1.0                              0         1.0   \n",
       "4                            1.0                              0         1.0   \n",
       "\n",
       "   liking  rating  rating_round  \n",
       "0     0.6     0.3           0.0  \n",
       "1     0.6     0.8           1.0  \n",
       "2     0.6     0.8           1.0  \n",
       "3     1.3     2.3           2.0  \n",
       "4     2.0     3.0           3.0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Unnamed: 0', 'session_id', 'track_id', 'skip_1', 'skip_2', 'skip_3',\n",
       "       'not_skipped', 'context_switch', 'hist_user_behavior_reason_start',\n",
       "       'hist_user_behavior_reason_end', 'hist_user_behavior_is_shuffle',\n",
       "       'skip_score', 'liking', 'rating', 'rating_round'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " # Import Decision Tree Classifier\n",
    "from sklearn.model_selection import train_test_split # Import train_test_split function\n",
    "from sklearn import metrics #Import scikit-learn metrics module for accuracy calculation\n",
    "df = df.dropna()\n",
    "len(df)\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.  -0.5  0.2  0.2  0.4  0. ]\n",
      " [ 0.   0.   0.2  0.2  0.4  0. ]\n",
      " [ 0.   0.   0.2  0.2  0.4  0. ]\n",
      " ...\n",
      " [-1.  -0.5  0.2  0.3  0.3  1. ]\n",
      " [ 0.  -0.5  0.2  0.3  0.3  1. ]\n",
      " [ 0.   0.   0.   1.   1.   0. ]]\n"
     ]
    }
   ],
   "source": [
    "X = df[['skip_1','skip_2','skip_3','hist_user_behavior_reason_start','hist_user_behavior_reason_end','hist_user_behavior_is_shuffle']]\n",
    "X = X.values\n",
    "print(X)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.]\n",
      " [1.]\n",
      " [1.]\n",
      " ...\n",
      " [0.]\n",
      " [1.]\n",
      " [3.]]\n"
     ]
    }
   ],
   "source": [
    "y = df[['rating_round']]\n",
    "y = y.values\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.15, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-1.  1. -0. ... -1.  2. -0.]\n",
      "Accuracy: 0.9711369160910446\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9711369160910446\n",
      "Precision: [1.         0.9866352  0.9489165  0.94160091 0.96657979 0.97332112]\n",
      "Recall: [0.99982182 0.99282737 0.92732212 0.94408425 0.96556245 0.99426255]\n",
      "F1 Score: [0.9999109  0.9897216  0.93799504 0.94284094 0.96607085 0.98368039]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\User\\Anaconda3\\lib\\site-packages\\sklearn\\ensemble\\forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n",
      "C:\\Users\\User\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:43: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-1.  1. -0. ... -1.  2. -0.]\n",
      "Accuracy: 0.9791956259846168\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9791956259846168\n",
      "Precision: [1.         0.99936674 0.94611586 0.95412492 0.9817878  0.98026316]\n",
      "Recall: [1.         0.99292175 0.97192117 0.9557326  0.96556245 0.99426255]\n",
      "F1 Score: [1.         0.99613382 0.95884492 0.95492808 0.97360753 0.98721323]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\User\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:724: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n",
      "C:\\Users\\User\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:193: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "### Decision Tree Classifier \n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "clf = DecisionTreeClassifier(max_depth=None, max_leaf_nodes=20)\n",
    "clf.fit(X_train, y_train)\n",
    "y_pred = clf.predict(X_test)\n",
    "print(y_pred)\n",
    "print(\"Accuracy:\",metrics.accuracy_score(y_test, y_pred))\n",
    "from sklearn.metrics import confusion_matrix\n",
    "cnf_matrix = metrics.confusion_matrix(y_test, y_pred)\n",
    "cnf_matrix\n",
    "class_names=[0,1] # name  of classes\n",
    "fig, ax = plt.subplots()\n",
    "tick_marks = np.arange(len(class_names))\n",
    "plt.xticks(tick_marks, class_names)\n",
    "plt.yticks(tick_marks, class_names)\n",
    "# create heatmap\n",
    "sns.heatmap(pd.DataFrame(cnf_matrix), annot=True, cmap=\"YlGnBu\" ,fmt='g')\n",
    "ax.xaxis.set_label_position(\"top\")\n",
    "plt.tight_layout()\n",
    "plt.title('Confusion matrix', y=1.1)\n",
    "plt.ylabel('Actual label')\n",
    "plt.xlabel('Predicted label')\n",
    "plt.show()\n",
    "dtree_accuracy = metrics.accuracy_score(y_test, y_pred)\n",
    "dtree_precision_score = metrics.precision_score(y_test, y_pred, average=None)\n",
    "dtree_recall_score = metrics.recall_score(y_test, y_pred, average=None)\n",
    "dtree_f1_score = metrics.f1_score(y_test, y_pred, average=None)\n",
    "print(\"Accuracy:\",dtree_accuracy)\n",
    "print(\"Precision:\",dtree_precision_score)\n",
    "print(\"Recall:\",dtree_recall_score)\n",
    "print(\"F1 Score:\", dtree_f1_score)\n",
    "plt.figure(figsize=(7,5))\n",
    "plt.hist(dtree_precision_score)\n",
    "plt.hist(dtree_recall_score)\n",
    "plt.hist(dtree_f1_score)\n",
    "plt.show()\n",
    "\n",
    "\n",
    "### RandomForestClassifier ####\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "clf2 = RandomForestClassifier(max_depth=None, max_leaf_nodes=20)\n",
    "clf2.fit(X_train, y_train)\n",
    "y_pred = clf2.predict(X_test)\n",
    "print(y_pred)\n",
    "print(\"Accuracy:\",metrics.accuracy_score(y_test, y_pred))\n",
    "from sklearn.metrics import confusion_matrix\n",
    "cnf_matrix = metrics.confusion_matrix(y_test, y_pred)\n",
    "cnf_matrix\n",
    "class_names=[0,1] # name  of classes\n",
    "fig, ax = plt.subplots()\n",
    "tick_marks = np.arange(len(class_names))\n",
    "plt.xticks(tick_marks, class_names)\n",
    "plt.yticks(tick_marks, class_names)\n",
    "# create heatmap\n",
    "sns.heatmap(pd.DataFrame(cnf_matrix), annot=True, cmap=\"YlGnBu\" ,fmt='g')\n",
    "ax.xaxis.set_label_position(\"top\")\n",
    "plt.tight_layout()\n",
    "plt.title('Confusion matrix', y=1.1)\n",
    "plt.ylabel('Actual label')\n",
    "plt.xlabel('Predicted label')\n",
    "plt.show()\n",
    "\n",
    "rftree_accuracy = metrics.accuracy_score(y_test, y_pred)\n",
    "rftree_precision_score = metrics.precision_score(y_test, y_pred, average=None)\n",
    "rftree_recall_score = metrics.recall_score(y_test, y_pred, average=None)\n",
    "rftree_f1_score = metrics.f1_score(y_test, y_pred, average=None)\n",
    "print(\"Accuracy:\",rftree_accuracy)\n",
    "print(\"Precision:\",rftree_precision_score)\n",
    "print(\"Recall:\",rftree_recall_score)\n",
    "print(\"F1 Score:\", rftree_f1_score)\n",
    "plt.figure(figsize=(7,5))\n",
    "plt.hist(rftree_precision_score)\n",
    "plt.hist(rftree_recall_score)\n",
    "plt.hist(rftree_f1_score)\n",
    "plt.show()\n",
    "\n",
    "\n",
    "#### SVM Classifier\n",
    "\n",
    "from sklearn import svm\n",
    "model = svm.SVC(kernel='rbf', C=1E6)\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "print(y_pred)\n",
    "from sklearn import metrics\n",
    "cnf_matrix = metrics.confusion_matrix(y_test, y_pred)\n",
    "cnf_matrix\n",
    "class_names=[0,1] # name  of classes\n",
    "fig, ax = plt.subplots()\n",
    "tick_marks = np.arange(len(class_names))\n",
    "plt.xticks(tick_marks, class_names)\n",
    "plt.yticks(tick_marks, class_names)\n",
    "# create heatmap\n",
    "sns.heatmap(pd.DataFrame(cnf_matrix), annot=True, cmap=\"YlGnBu\" ,fmt='g')\n",
    "ax.xaxis.set_label_position(\"top\")\n",
    "plt.tight_layout()\n",
    "plt.title('Confusion matrix', y=1.1)\n",
    "plt.ylabel('Actual label')\n",
    "plt.xlabel('Predicted label')\n",
    "plt.show()\n",
    "\n",
    "svm_accuracy = metrics.accuracy_score(y_test, y_pred)\n",
    "svm_precision_score = metrics.precision_score(y_test, y_pred, average=None)\n",
    "svm_recall_score = metrics.recall_score(y_test, y_pred, average=None)\n",
    "svm_f1_score = metrics.f1_score(y_test, y_pred, average=None)\n",
    "print(\"Accuracy:\",svm_accuracy)\n",
    "print(\"Precision:\",svm_precision_score)\n",
    "print(\"Recall:\",svm_recall_score)\n",
    "print(\"F1 Score:\", svm_f1_score)\n",
    "plt.figure(figsize=(7,5))\n",
    "plt.hist(svm_precision_score)\n",
    "plt.hist(svm_recall_score)\n",
    "plt.hist(svm_f1_score)\n",
    "plt.show()\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_accuracy = [dtree_accuracy,rftree_accuracy,svm_accuracy]\n",
    "model_accuracy = pd.DataFrame(model_accuracy)\n",
    "sns.heatmap(model_accuracy)\n",
    "plt.show()\n",
    "\n",
    "model_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_precision = [dtree_precision_score,rftree_precision_score,svm_precision_score]\n",
    "model_precision = pd.DataFrame(model_precision)\n",
    "sns.heatmap(model_precision)\n",
    "plt.show()\n",
    "\n",
    "model_precision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_recall = [dtree_recall_score,rftree_recall_score,svm_recall_score]\n",
    "model_recall = pd.DataFrame(model_recall)\n",
    "sns.heatmap(model_recall)\n",
    "plt.show()\n",
    "\n",
    "model_recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import svm\n",
    "model = svm.SVC(kernel='rbf', C=1E6)\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = pd.DataFrame(y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['rating_round'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#### Creating a Full y_pred list with based in SVM Classifier ####\n",
    "from sklearn import svm\n",
    "model = svm.SVC(kernel='rbf', C=1E6)\n",
    "model.fit(X,y)\n",
    "svm_pred = model.predict(X)\n",
    "svm_pred = pd.DataFrame(svm_pred)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['svm_pred'] = svm_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['svm_pred'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### clustering of user ID based on the predicted ratings\n",
    "\n",
    "rating_1_uid = df.query('svm_pred==-1')['session_id']\n",
    "rating_1_uid = pd.DataFrame(rating_1_uid)\n",
    "\n",
    "rating_2_uid = df.query('svm_pred==0')['session_id']\n",
    "rating_2_uid = pd.DataFrame(rating_2_uid)\n",
    "\n",
    "rating_3_uid = df.query('svm_pred==1')['session_id']\n",
    "rating_3_uid = pd.DataFrame(rating_3_uid)\n",
    "\n",
    "rating_4_uid = df.query('svm_pred==2')['session_id']\n",
    "rating_4_uid = pd.DataFrame(rating_4_uid)\n",
    "\n",
    "rating_5_uid = df.query('svm_pred==3')['session_id']\n",
    "rating_5_uid = pd.DataFrame(rating_5_uid)\n",
    "\n",
    "rating_6_uid = df.query('svm_pred==4')['session_id']\n",
    "rating_6_uid = pd.DataFrame(rating_6_uid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rating_1_uid.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rating_2_uid.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat = pd.read_csv(r'dat.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat['hist_user_behavior_is_shuffle'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat['svm_rating'] = svm_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat_arr = dat.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "z = np.random.rand(15)\n",
    "z=z*z\n",
    "plt.figure(figsize=(10,8)) \n",
    "\n",
    "# Change color with c and alpha. I map the color to the X axis value.\n",
    "plt.scatter(dat['release_year'], dat['Unnamed: 0'], s=dat['hour_of_day']*50, c = dat['svm_rating'] , alpha=0.4, linewidth=2)\n",
    "\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat['hour_of_day'].describe()\n",
    "plt.hist(dat['acousticness'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "dat['session_length'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "z = np.random.rand(15)\n",
    "z=z*z\n",
    "plt.figure(figsize=(10, 8)) \n",
    "\n",
    "# Change color with c and alpha. I map the color to the X axis value.\n",
    "plt.scatter(dat['acousticness'], dat['Unnamed: 0'], s=dat['hour_of_day']*10, c = dat['svm_rating'] , alpha=1, linewidth=2)\n",
    "\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rating_1_uid.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = rating_1_uid.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = [n]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "type(n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n2 = pd.DataFrame(n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n2 = n2.transpose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n2 = list(n2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8)) \n",
    "\n",
    "\n",
    "\n",
    "plt.scatter(dat['beat_strength'], dat['bounciness'],cmap=plt.get_cmap(\"jet\"),\n",
    "            c = dat['svm_rating'],s = dat['hour_of_day'])\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "g = sns.FacetGrid(dat, col=\"svm_rating\",  hue=\"hour_of_day\")\n",
    "g = (g.map(plt.scatter, \"beat_strength\", \"bounciness\", edgecolor=\"w\").add_legend())\n",
    "g.fig.set_size_inches(10,5)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "g = sns.FacetGrid(dat, col=\"svm_rating\",  hue=\"hist_user_behavior_is_shuffle\")\n",
    "g = (g.map(plt.scatter, \"beat_strength\", \"bounciness\", edgecolor=\"w\").add_legend())\n",
    "plt.colorbar()\n",
    "g.fig.set_size_inches(10,5)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "g = sns.FacetGrid(dat, col=\"hist_user_behavior_is_shuffle\",  hue=\"svm_rating\")\n",
    "g = (g.map(plt.scatter, \"beat_strength\", \"bounciness\", edgecolor=\"w\").add_legend())\n",
    "g.fig.set_size_inches(12,7)\n",
    "plt.colorbar()\n",
    "\n",
    "plt.show()\n",
    "\n",
    "\n",
    "g1 = sns.FacetGrid(dat, col=\"hist_user_behavior_reason_start\",  hue=\"svm_rating\")\n",
    "g1 = (g1.map(plt.scatter, \"beat_strength\", \"bounciness\", edgecolor=\"w\").add_legend())\n",
    "g1.fig.set_size_inches(12,5)\n",
    "plt.colorbar()\n",
    "plt.show()\n",
    "\n",
    " \n",
    "g2 = sns.FacetGrid(dat, col=\"hist_user_behavior_reason_end\",  hue=\"svm_rating\")\n",
    "g2 = (g2.map(plt.scatter, \"beat_strength\", \"bounciness\", edgecolor=\"w\").add_legend())\n",
    "g2.fig.set_size_inches(12,5)\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "dat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat['session_length'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "############# Clustering ############"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "g2 = sns.FacetGrid(dat, col=\"svm_rating\",  hue=\"hist_user_behavior_is_shuffle\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = dat['Unnamed: 0'].values\n",
    "y = dat['svm_rating'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##### make Blobs #######"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import LabelEncoder\n",
    "from sklearn import preprocessing\n",
    "#creating labelEncoder\n",
    "le = preprocessing.LabelEncoder()\n",
    "G = le.fit_transform(dat['Unnamed: 0'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(dat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_blobs\n",
    "m,n = make_blobs(G[500],centers=3,random_state=0, cluster_std=0.65)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Kmean.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.scatter(-1.52711596,  2.93873372,c = 'g', marker = 's')\n",
    "plt.scatter(1.95873501,  0.80271418,c = 'black', marker = 's')\n",
    "plt.scatter(0.90206723,  4.36135952,c = 'b', marker = 's')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "gth = df[['Unnamed: 0','rating_round']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "g = sns.FacetGrid(dat, col=\"hist_user_behavior_is_shuffle\",  hue=\"svm_rating\")\n",
    "g = (g.map(plt.scatter, \"beat_strength\", \"bounciness\", edgecolor=\"w\").add_legend())\n",
    "g.fig.set_size_inches(12,8)\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data3 = dat[['beat_strength','bounciness']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data3_arr = data3.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data3_arr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G2 = le.fit_transform(data3['beat_strength'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "m,n = make_blobs(G2[500],centers=3,random_state=0, cluster_std=0.65)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Kmean.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.scatter(2.05314174,  0.89176704,c = 'g', marker = 's')\n",
    "plt.scatter(1.95873501,  0.80271418,c = 'black', marker = 's')\n",
    "plt.scatter(-1.53745756,  2.91305784,c = 'b', marker = 's')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(data3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G3 = le.fit_transform(data3['bounciness'])\n",
    "m,n = make_blobs(G3[167879],centers=3,random_state=0, cluster_std=0.45)\n",
    "plt.figure(figsize=(20, 18)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.legend()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)\n",
    "Kmean.cluster_centers_\n",
    "plt.scatter( 0.9738295 ,  4.30430402,c = 'grey', marker = 's')\n",
    "plt.scatter(2.05084236,  0.90036442,c = 'black', marker = 's')\n",
    "plt.scatter(-1.52585357,  2.91613035,c = 'blue', marker = 's')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import kde\n",
    "\n",
    "fig, axes = plt.subplots(ncols=2, nrows=1, figsize=(21, 5))\n",
    " \n",
    "# Everything sarts with a Scatterplot\n",
    "plt.figure(figsize=(20, 15))\n",
    "axes[0].set_title('Scatterplot')\n",
    "axes[0].plot(m[:, 0], m[:, 1])\n",
    "\n",
    "# As you can see there is a lot of overplottin here!\n",
    " \n",
    "# Thus we can cut the plotting window in several hexbins\n",
    "nbins = 20\n",
    "plt.figure(figsize=(20, 15))\n",
    "axes[1].set_title('Hexbin')\n",
    "axes[1].hexbin(m[:, 0], m[:, 1], gridsize=nbins, cmap=plt.cm.BuGn_r)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G4 = le.fit_transform(dat['svm_rating'])\n",
    "m,n = make_blobs(G4[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(20, 18)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='viridis')\n",
    "\n",
    "\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)\n",
    "Kmean.cluster_centers_\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G4 = le.fit_transform(dat['session_position'])\n",
    "m,n = make_blobs(G4[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)\n",
    "Kmean.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G4 = le.fit_transform(dat['session_position'])\n",
    "m,n = make_blobs(G4[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)\n",
    "print(Kmean.cluster_centers_)\n",
    "\n",
    "G5 = le.fit_transform(dat['session_length'])\n",
    "m,n = make_blobs(G5[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m[:, 0], m[:, 1], c=n, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m)\n",
    "print(Kmean.cluster_centers_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G6 = le.fit_transform(dat['session_id'])\n",
    "m2,n2 = make_blobs(G6[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m2[:, 0], m2[:, 1], c=n2, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m2)\n",
    "print(Kmean.cluster_centers_)\n",
    "\n",
    "G7 = le.fit_transform(dat['hist_user_behavior_reason_start'])\n",
    "m1,n1 = make_blobs(G7[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m1[:, 0], m1[:, 1], c=n1, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean2 = KMeans(n_clusters=3)\n",
    "Kmean2.fit(m1)\n",
    "print(Kmean2.cluster_centers_)\n",
    "\n",
    "G8 = le.fit_transform(dat['hist_user_behavior_reason_end'])\n",
    "m3,n3 = make_blobs(G8[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m3[:, 0], m3[:, 1], c=n3, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean3 = KMeans(n_clusters=3)\n",
    "Kmean3.fit(m3)\n",
    "print(Kmean3.cluster_centers_)\n",
    "\n",
    "G9 = le.fit_transform(dat['session_length'])\n",
    "m9,n9 = make_blobs(G9[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m9[:, 0], m9[:, 1], c=n9, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean9 = KMeans(n_clusters=3)\n",
    "Kmean9.fit(m)\n",
    "print(Kmean9.cluster_centers_)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sid_cluster = Kmean.cluster_centers_\n",
    "start = Kmean2.cluster_centers_\n",
    "end = Kmean3.cluster_centers_\n",
    "session_length = Kmean9.cluster_centers_\n",
    "#shuffle = Kmean4.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sid_cluster = pd.DataFrame(sid_cluster)\n",
    "start = pd.DataFrame(start)\n",
    "end = pd.DataFrame(end)\n",
    "sess = pd.DataFrame(session_length)\n",
    "#shuffle = pd.DataFrame(shuffle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sid_cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "start"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8))\n",
    "plt.scatter(sid_cluster[0],sid_cluster[1],c = 'g', marker = 's')\n",
    "plt.scatter(start[0],start[1],c = 'blue', marker = 's')\n",
    "plt.scatter(end[0],end[1],c = 'red', marker = 's')\n",
    "plt.scatter(sess[0],sess[1],c = 'orange', marker = 's')\n",
    "color = ['session_id','start_session','end_session','session_length']\n",
    "plt.legend(color)\n",
    "plt.xticks()\n",
    "plt.yticks()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "g2 = sid_cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "start"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d2 = pd.DataFrame((sid_cluster[0]-start[0]).astype('float')**2 + (sid_cluster[1]-start[1]).astype('float')**2)\n",
    "d3 = pd.DataFrame((sid_cluster[0]-end[0]).astype('float')**2 + (sid_cluster[1]-end[1]).astype('float')**2)\n",
    "d4 = pd.DataFrame((sid_cluster[0]-sess[0]).astype('float')**2 + (sid_cluster[1]-sess[1]).astype('float')**2)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(d2)\n",
    "print(d3)\n",
    "print(d4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d22 = d2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d22['end'] = d3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d22['sess'] = d4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d22.rename(columns = {0:'start'}, inplace = True) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 7))\n",
    "d22.boxplot(column=['start','end','sess'],patch_artist='True',vert=True,showmeans=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#d2 = math.sqrt( ((sid_cluster[0]-start[0])**2)+((sid_cluster[1]-start[1])**2) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dat.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "G6 = le.fit_transform(dat['session_id'])\n",
    "m2,n2 = make_blobs(G6[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m2[:, 0], m2[:, 1], c=n2, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean = KMeans(n_clusters=3)\n",
    "Kmean.fit(m2)\n",
    "print(Kmean.cluster_centers_)\n",
    "\n",
    "\n",
    "G11 = le.fit_transform(dat['beat_strength'])\n",
    "m11,n11 = make_blobs(G11[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m11[:, 0], m11[:, 1], c=n11, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean11 = KMeans(n_clusters=3)\n",
    "Kmean11.fit(m11)\n",
    "print(Kmean11.cluster_centers_)\n",
    "\n",
    "G12 = le.fit_transform(dat['acousticness'])\n",
    "m12,n12 = make_blobs(G12[167879],centers=3,random_state=0, cluster_std=0.55)\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m12[:, 0], m12[:, 1], c=n12, s=30, cmap='autumn')\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean12 = KMeans(n_clusters=3)\n",
    "Kmean12.fit(m12)\n",
    "print(Kmean12.cluster_centers_)\n",
    "\n",
    "G13 = le.fit_transform(dat['bounciness'])\n",
    "m13,n13 = make_blobs(G13[167879],centers=3,random_state=0, cluster_std=0.45) \n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.scatter(m13[:, 0], m13[:, 1], c=n13, s=30, cmap='autumn')\n",
    "\n",
    "plt.show()\n",
    "from sklearn.cluster import KMeans\n",
    "Kmean13 = KMeans(n_clusters=3)\n",
    "Kmean13.fit(m13)\n",
    "Kmean13.cluster_centers_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "beat = pd.DataFrame(Kmean11.cluster_centers_)\n",
    "ac = pd.DataFrame(Kmean12.cluster_centers_)\n",
    "bounce = pd.DataFrame(Kmean13.cluster_centers_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sid_cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "beat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ac"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8))\n",
    "plt.scatter(sid_cluster[0],sid_cluster[1],c = 'g', marker = 's')\n",
    "plt.scatter(beat[0],beat[1],c = 'blue', marker = 's')\n",
    "plt.scatter(ac[0],ac[1],c = 'r', marker = 's')\n",
    "plt.scatter(bounce[0],bounce[1],c = 'grey', marker = 's')\n",
    "color = ['session_id','beat','ac','bounce']\n",
    "plt.legend(color)\n",
    "plt.xticks()\n",
    "plt.yticks()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d5 = pd.DataFrame((sid_cluster[0]-beat[0]).astype('float')**2 + (sid_cluster[1]-beat[1]).astype('float')**2)\n",
    "d6 = pd.DataFrame((sid_cluster[0]-ac[0]).astype('float')**2 + (sid_cluster[1]-ac[1]).astype('float')**2)\n",
    "d7 = pd.DataFrame((sid_cluster[0]-bounce[0]).astype('float')**2 + (sid_cluster[1]-bounce[1]).astype('float')**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d23 = d5\n",
    "d23['ac'] = d6\n",
    "d23['bounce'] = d7\n",
    "d23.rename(columns = {0:'beat'}, inplace = True) \n",
    "d23"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 7))\n",
    "d23.boxplot(column=['beat','ac','bounce'],patch_artist='True',vert=True,showmeans=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 8))\n",
    "plt.scatter(sid_cluster[0],sid_cluster[1],c = 'g', marker = 's')\n",
    "plt.scatter(beat[0],beat[1],c = 'blue', marker = 's')\n",
    "plt.scatter(ac[0],ac[1],c = 'r', marker = 's')\n",
    "plt.scatter(bounce[0],bounce[1],c = 'grey', marker = 's')\n",
    "plt.scatter(start[0],start[1],c = 'orange', marker = 's')\n",
    "plt.scatter(end[0],end[1],c = 'black', marker = 's')\n",
    "plt.scatter(sess[0],sess[1],c = 'pink', marker = 's')\n",
    "color = ['session_id','beat','ac','bounce','start','end','sess']\n",
    "plt.legend(color)\n",
    "plt.xticks()\n",
    "plt.yticks()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d45 = d22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d45"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "d45['ac'] = d6\n",
    "d45['bounce'] = d7\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d45['beat'] = d23['beat']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d45"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(10, 7))\n",
    "d45.boxplot(column=['start','end','sess','ac','bounce','beat'],patch_artist='True',vert=True,showmeans=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}